documentv>
TRANSCRIPT
![Page 1: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/1.jpg)
\
. '
'f
W:' .uAM: B. ROBINSON , ~ CoDJlllissioner
ERSKIND DeRA.l."'vIUS Deputy Commissioner
V><THE IMPACT OF AN ORGANIZED VARSITY SPORTS PROGRAM ~
iN THE PENNSYLVANIA BUREAU OF CORRECTION AS IT RELATES TO INMATE SELF-ESTEEM
Gerard N. Massaro, Ph.D.
Director
Planning & Research Division
by
W. Scott Thornsley
Planner/Evaluator
If you have issues viewing or accessing this file, please contact us at NCJRS.gov.
![Page 2: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/2.jpg)
, ..
.. ABSTRACT
Thorns I ey, W. Scott, The Impact of an Organized Vars1:ty Sports Program in the PennsyZvania Bureau of Correction as it ReZates to SeZfEsteem. March, 1979, Pennsylvania Bl:J/eau of Correction r• Camp Hill, Pennsylvania.
Purpose:
The objective of this study was to determine if inmate participation in
the Pennsylvania Bureau of Correction's organized varsity sports program is
associated with inmate self-esteem.
Methods:
The methods used in this study were: (1) distribution of Coopersmith's
(1961) Self-Esteem Inventory (SEI) to randomly selected varsity sports partici
pants (N=lOO) and a matched p~ir sampling of non-participants (N=IOO), based
on specific variables, and (2) sampling of March and April, 1978 inmate
receptions so they may be subsequently placed in one of four groups found in
the Randomized Solomon Four-Group Des!gn, i.e. Phase I I. (April receptions
were identified as groups 1 and 2 since they were pretested with the SEI, March
receptions were identified as groups 3 and 4. All four groups were posttested.),
(3) conducting a computer analysis of this data to determine frequency distribu-
tions for five variables, and (4) considering all results with a probabil ity
of .05 or less as significant.
Find i ngs ::
1. The study indicates that no significant difference exists between sports
participants and non-participants as it relates to, self-esteem, as found in Phase I.
2. There was no statistical significance found between the mean scores
of the SEI of the participants and non-participants while control 1 ing for the
effects of pretesting and other outside variables.
![Page 3: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/3.jpg)
..
..
3. There was negl igible association found between the SEI scores of
sports participants while cantrall ing for race,
4. There was negl igible association found between the SEI scores of
sports participants while contraIl ing for race.
5. There was negl igible association found between the SEI scores of
non-participants while cantrall ing for race.
6. While not significant at the probabil ity level of .05, the SEI
scores generally increase after initial incarceration for both participants
and non-participants.
7. While not significant at the .05 probabil ity level, individuals
who become involved with the organized varsity sports program generally have
a higher level of self-esteem prior to their involvement.
8. The self-esteem of the non-participant (while still mathmatically
lower than that of the participant) generally increases more than that of the
participantls increase of self-esteem.
9. A statistically significant dIfference (t=I,8080, df=176, p=.05)
was found between the pre and posttests of April IS non-participants. The
groupls mean SEI score increased from 1.687 to'18.147, a net mean increase
of 1.27.
10. Generally speaking, inmates who were later to be identified as
sports participants displayed a mathmatically higher mean SEI score than
non-participants (18.500 vs. 16.876 respectively). This mathmatically higher
mean SEI score was also true in reviewing the posttest scores of the two
groups (19.500 vs. 18.146, respectively).
While the non~participant has displayed a mathmatical1y lower mean SEI
score, the non-participant has increased his mean SEI scores at a faster rate
than has the sports participant. This leads the researcher to bel ieve that
iii
![Page 4: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/4.jpg)
sooner or later the mean SEI Scores found between the four groups will
.. g radua 11 y merge.
11. Generally speaking,the inmates who participate in the Bureau'z
organized varsity sports program tend to display a mathmatica11y higher
mean SEI Score than the inmate non-participant prior to that involvement.
While this difference is not statistically significant, it allows the reader
to assume that inmates who participate in the organized varsity sports program
have a sound level of self-esteem prior to that athletic involvement, not
because of that participation. Stated simply, organized sports program will
attract inmates who already have an adequate self-esteem.
iv
![Page 5: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/5.jpg)
..
TABLE OF CONTENTS
PAGE
ABSTRACT •.•... » ••••••• " •• t ••••••••••••••••••••••• , , ••••• " • " ••••••••• ,. i i
LIST OF TABLES .•.•.•.•.••...• " ............. , ..••..•..•.••••. t •••••••••• vi i
LIST OF FIGURES ..•......•. '! il .................................... i> ••••• vi i i
CHAPTER
I. I NT-RODUCT ION ....•.•........ , .. ' ............. " .. , .. , ......... . Research ,Issue ••.......................... III ••••••••••••••••• 2
Problem Statement ......................•.....•.............. 4
I I. REV I EW OF THE LITERATURE .......••.......•......... ,........... 5
III. METHODOLOGY .•...... , .•..•.•.•......• \I ••••••••••••••••••••••••• 8
Survey Instrument •.. . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . 8
Scoring of the SE I •. 8
Research Methods Uti1 ized ....... , ••.......•................. 9
Chosen Level of Significance. .................... , ..... , .... 11
I nst rument Re1iabi1 ity. 11
Instrument Va 1 i d i ty. 12
Analysis of Data ............................................ 12
Study Sites. ' ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
IV. ANALY.SIS OF DATA 13
Phase Findings .• 13
Phase I I Findings. " ••••••••••••••••••••• I," •••••••••••••• 18
Group Analysis ........•..... t ••••• " •••••••••••••••••••• t, •••• 29
V. CONCLUSIONS •..••.•. 1I ........ t ...... , ••• l1li ••••••••••••• , •••••••••••• 37
![Page 6: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/6.jpg)
FOOTNOTES .. , .....•..•... , .......•.....•.....•..••........•...• ,. ... PAGE 39
BIBLIOGRAPHy...................................................... 41
APPENDiCES •.......•.............•........•. " .......... 1: ••••••••••• 42
Appendix A. The Self-Esteem Inventory ........................ . 43
vi
i.,','
~
![Page 7: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/7.jpg)
LIST OF TABLES
TABLE PAGE
1. Self-Esteem Inventory (SE I) Responses for Phase I .................•.. 14
2. SEI Responses for Sports Part icipants ........................•.. 41 •••• 15
3. SEI Responses for Non-Part i c i pants ..................•................ 16
4. SEI Responses of Phase II's Group #1 (Pretest) ....................... 23
5. SEI Responses of Phase II's Group #1 (Posttest) .................... II' • 24
6. SEI Responses of Phase Ills Group #2 (Pretest) ....................... 25
7. SEI Responses of Phase II IS Group #2 (Posttest) ...................... 26
8. SEI Responses of Phase II's Group #3 (Posttest) ...................... 27
9. SEI Responses of Phase II's Group #4 (Posttest) ...................... 28
10. Breakdown of the SEI by Race and Group ................................. 30
11. SEI Responses of Posttests in Phase II for Paft i c i pants .............. 32
12. SEI Responses of Post tests in Phase II for Non-Participants ...•...... 33
13. SEI Responses of Post tests in Phase II for Whites .................... 35
14. SEI Responses of Posttests in Phase II for Blacks ..................... 36
· tl l'
;j .• "
"
--~
![Page 8: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/8.jpg)
FIGURE
1.
2.
LIST OF FIGURES
Randomized Solomon Four Group Design .........................•...
Randomized Solomon Four Group Design For Phase I I Receptions ...•.
PAGE
20
21
J
![Page 9: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/9.jpg)
'.,
CHAPTER 1
INTRODUCTION
The purpose of this chapter is to present the reader with a discussion
of the issue and problem statement so as to enable the reader to comprehend
the purpose of this study and its potential importance to the correctional
administrator who is always looking for new methods to introduce rehabi1 itative
programs.
The fact that sports is an important part of prison 1 ife was brought
home to mill ions of American television viewers on Saturday February 10, 1979
as they viewed Rahway State Prison inmate James Scott defeating Richie Kates
in a nationally televised 1 ight-heavyweight bout fought inside the prison on
"CBSls Sports Spectacu1ar. 11
Scott, who is on his way to gain a bout with the current 1 ight-
heavyweight champion, is an articulate inmate who has devoted his prison
1 ife to boxing. To Scott, boxing " ... makes 1 ife behind these walls just
al ittle bit better.'1
Scott is not the first prison inmate who has gained national prominence 'ti""· ~ ..
while incarcerated due to his boxing skills. Boxing in particular seems to
draw the majority of interest in prison sports. Perhaps that is why so
many state prison systems already offer or are considering instituting an
expanded sports program. Inmates like James Scott and Bobby Lee Hunter (an
inmate from Sourth Carol inals Manning Correctional Institution, who achieved
national pub1 icity during 1971-72 when he was striving to make the U.S.
Olympic team for the 1972 Munich 01ymics) appear to rise to the top in national
athletic competition in spite of~ not because of, their prisonls organized
sports programs. ~
fj' r
I I J!
----------------------------------------------------------------------------------- ;
![Page 10: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/10.jpg)
..
--~---------------------------------------------------.----------~--~~
The P~nnsylvania Bureau of Correction has done much to improve its
organized sports program within the past four years. Sports programming
in the Bureau of Correction has evolved from a mediocre intramural program
into a highly organized varsity sports program which sponsors regularly
scheduled inter-institutional athletic events in the sports of football,
basketball, softball and boxing.
Research Issue
Whether or not to fund an organized varsity sports program in prisons
2
is certainly not the most particularly pressing problem the state correctional
administrator faces today; however, it is an issue they must be prepared to
respond to, for prison inmates as well as civ'il 1 ibertarians are demanding3 not
requesting, that increased money, personnel and institutional time be available
for recreation or sporting events.
The demand for increased recreational and athletic activities in prisons
is not a new concept, though it is most certainly gaining momentum from society
at large. A recent ABC news commentary focused on the leisure activity pf the
American citizen, and was surprised to find that now individuals view recreation
as an inherent right and not as a luxury only the well-to-do could once afford.
This inherent right to recreation within a prison system was dramatically brought
to light during the 1971 Attica Prison uprising. This demand for increased
recr'f.~ational activities was viewed as so important by the Attica insurgents
that they included it as one of their famous "Fifteen Practical Proposals."
This proposal was une of "twenty-eight points" that then New York State Commis
sioner of Correction Russell G. Oswald agreed to accept during the uprising.
1
![Page 11: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/11.jpg)
.. Today most state prison systems have some sort of intramural sports
program operating within their prisons, but few have let intramural competi-
tion evolve beyond much more than a form of in~house institutional recreation.
To the state prisons that use sports activity as recreation -- inmate
participation then becomes just that ~- recreation. However, when a state
prison system decides to let the concept of organized varsity sports ON a
state-wide basis develop, it begins to take on an added characteristic.
Inmate participation in an organized sports program which operat6s on a
state-wide systeh; then can begin to develclp into an unintentionally disguised
treatment program designed to address the complex personal problems of anxiety,
boredom, frustration, lack of achievement, self-confidence and self-esteem.
Involvement in a prison's organized sports program has been encouraged
by numerous correctional administrators, both past and present. Former Sing
Sing Warden Lewis E. Lawes noted in his TWenty Thousand Years at Sing Sing
(1932) that
... in this period of play and healthful recreation there is no time or incll~~tion to think of self. The body is too a:tiye, the mind intent on immediate contests and the arm to win.
Many more correctional administrators are becoming aware of the hereto-
fore unexplored possibil ities of recreation as treatment.
Under the guise of recreation, personal ity problems may be approached by the staff through the "back door" because they permit informal staff-inmate relationships. The shy, inadequate person may be drawn out, and the bully may be guided into more acceptable ways of relating himself with others. 2
Others see recreation in a different I ight, while at the same time recog-
nizing its importance. Tappan's Crime~ Justice and Correction (1960) states:
Recreational activities are important to inmate morale and are significant, therefore, to effective insitutional discipl ine. Furthermore, it is increasing apparent that expanding leisure time in the free community calls increasingly for the development of habits, interests, and enjoyments to which, thus far, our society has paid little heed.3
3
I I
![Page 12: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/12.jpg)
.-The issue of providing an organized varsity sports program on a state-
wide basis in a state prison system will entail a great deal of reflection
and commitment on the part of the correctional administrator. The question
of whether or not to commit time, money and personnel to a program such as
this will hinge primarily on whether or not the correctional administrator
is will ing to admit that involvement in organized sports can have a posi-
tive effect on the entire committed inmate population.
Problem Statement
The research problem that this study will address is the effect of a
prison1s organized sports program upon its participants. While this study
(nor any other) cannot control for all variables, completion of the study
will allow the researcher to state that involvement in an organized sports
program while incarcerated may be associated with an inmate1s level of self-, esteem.
Specifically, then, the operational problem that this study will examine
is to investigate the hypothesis that inmates who participate in one of the
Bureau1s four organized sports programs (football, basketbBII, softball and
boxing) display a higher level of self-esteem as opposed to the general inmate
population. Therefore, the hypothesis that this stuU~ will util ize is the
following:
Those inmates who participate in organized prison sports programs will display a higher level of self-esteem than those inmates who do not participate.
4
![Page 13: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/13.jpg)
.-..
CHAPTER 2
REVIEW OF THE LITERATURE
Just as organized prison varsity sports programs are a relatively recent
innovation in prison programming, so too are the evaluations of. those programs.
The only major work to date which has addressed the association between
involvement in an institutional sports program and institutional ~djustment,
recidivism, etc., was conducted by Medve in 1961. Unfortunately Medve's work,
liThe Rehabilitative Aspects of Team Sports in a Reformatory," which was pub-
1 ished in the Journal of Correctional Education did not contain a control
group, thus negating the possibi1 ity of positively stating that institutional
sports do have a significant impact upon the incarcerated. However, Medve's
study did induce Martinson to state in his landmark work, The Effectiveness
of Correctional Treatment: A Survey of Treatment Evaluation Studies to state
... Medve's findings suggest, at least, that meaningful institutional programming may have a~ impact upon both institutional behavior and post~re1ease behavior.
Specifically, Medve's study examined the association between participation
in intramural and varsity level sports and recidivism and parole violation
for y.oung males confined to a reformatory,
In this study (Medve, 1961), the author discovered that those young male reformatory inmates who participated moderately or on very active levels (intramural or varsIty) violated parole during a one-year follow-up less frequently than those who were inactive in sports. The absence: of a control group in this study, however, makes these findings tentative since it may be that young men who chose to participate in sports may be less 1 ikely to violate parole than thos~ who do not so chose. 5
![Page 14: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/14.jpg)
.-Martinson's mention of Medve's work enables the reader to real ize just
hoW neglected this area of correctional evaluation has been, as the purpose
of Martinson's work was to ...
.. . provide a comprehensive compilation and analysis of local, national, and foreign research studies that have been and are currently being conducted to evaluate the treatment of criminal and juvenile offenders (from January 1, 1945 to December 31, 1967). The basic questions asked were, What treatment methods have been administered to criminal offenders? What can be said of their effectiveness in changing the offender or in reducing recidivism7 6
This does not imply that no studies have been conducted which
examine the association between athletics and del inquency . Schaefer's
"Participation in Interscholastic Athletics and Delinquency: A Prel im-
inary Study," which appeared in a 1969 issue of SociaZ ProbZems makes
strong suggestions as to the association between participation in int~r-
scholastic athletics and its deterring effect on del inquency. While
Schaefer's article does not address institutional sports participation,
his predictions are relevant to Medve's findings and to sports involve-
ment in general. Schaefer states that ...
... to predict that other things being equal, participation in interscholastic athletics will have a deterring influence on del inquency. Stated in testable terms, athletes will have a lower delinquency rate than non-athletes, other things being equa 1.7
This deterring influence ment}oned by Schaefer Is important to
correctional administrators who plan activities ,f~r those incarcerated.
Even if their only goal is to reduce the amount of institutional tension
and turmoil, sports programming is worthy of serious expansion. What is
true in society's ghettos is equally true in their prisons.
6
1
I',;,', i , -' \ I i
, "
![Page 15: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/15.jpg)
-'"C~"~".....,--"..,,.,,,-.,----,,,.-.,---.,-----,-___________________________ _
... Theorists contend that del inquency often arises out of sheer boredom ... Clearly, athletes are less 1 ikely to be bored and thereby suspectible to del inquency than are comparable nonathletes, since sports take so much time ... Even when not directly occupied in practice or competition, the athlete's psychic gmerges and loyalties are still directed in a conforming direction.
Nolan's (1954) article about. the association between juvenile
delinquency and athletics also addresses the question of preventative
del inquency. Nolan was of the opinion that athletic participation enabled
the individual to conform to socially accepted standards by stating:
Play becomes a natural outlet for the competitive instinct, and guided play becomes a way of learning about I ife ... The rules of social behavior, the "shall" and "shall not," become gradually a part of his whole personal ity and makeup and are carried over, naturally and easily, into the world that exists outside sports, the world that will soon absorb him.9
While participation in a prison-community's organized varsity sports
program may not be the answer for the entire inmate population, it certainly
is for a considerable percent. Participation in organized athletic competition
is probably one of the few areas where participants and spectators alike can
escape from the tedious existence of day to day incarceration. In Clemmer's
The Prison Community (1940), the effect of the institution was regarded as
harmful to the rehabilitation of the inmate. It is the hope of many state
prison administrators that by enlarging their recreational programs, especially
their sports programs, that these effects will not be as severe.
In the words of Rahway State Prison inmate James Scott, future contender
for the light-heavyweight championship of the world, those " ... left jabs and
right hooks keep me going."
7
,----------------------------------------------------------------------------------~---------~
![Page 16: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/16.jpg)
CHAPTER III
METHODOLOGY
The method chosen to evaluate the Pennsylvania Bureau of Correction's
organized sports program will be to examine the inmate athlete's (hereinafter
participant) level of self-esteem, and to see if participation in anyone of
the bureau's organized varsity sports programs (basketball, softball, football
and boxing) could be one of the factors responsible for higher level of se1f-
esteem as compared against non-participants.
Survey Instrument
The survey instrument that this study will uti1 ize is the Self-Esteem
Inventory (SEI), developed by Stanley Coopersmith in 1967 (see 8ppendix A).
The original SEI was composed of 50 items and primarily drawn from the work
of Rogers and Dymond (1954). The 50 item SEI was then reduced to 25 items
so as to avail itself for use by subjects of all ages. In comparison to the
long 50 item form, the short 25 item form yielded a correlation of over .95
when tested.
The statements are answered either "like mell or "unl ike me" and are
multi-dimensional in nature. The dimensional aspects of the SEI yield four
distinct areas of concern. They are as follows: (1) self-derogation,
(2) leadership-popularity, (3) family-parents, and (4) assertiveness-anxietys
with II ••• the fami ly-parents 10 factor being the most stable and unambiguous. 1I
Scoring of the SEI
Scorin~ of the SEI was done manually with the use of a' templet which
identified the high-esteem responses for each of the 25 items. Thus the
range of the inventory is 00 (lowest_self-esteem possible) to 25 (highest
self-esteem possible).
![Page 17: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/17.jpg)
Research Models Util ized
This evaluation util ized two research designs over the project's 12 month
period. The use of two research designs was not an attempt to elaborate on the
hypothesis, but rather to compl iment each other with respect to instrument val i-
dlty. The use of two separate but related research designs also made necessary
that this study be comprised of two distinct phases of data collection.
The first phase of the evaluation util ized a quasi-experimental design,
which is methodologically found between the level of d truly controlled experi-
mental design and a nonexperimental design. This type of design is enjoying
popularity in research methodology as they " ... are intended for situations in 11
which true experimentation is not feasible of desireable •.. "
Rigorously controlled experiments are possibJ~ in correction more often than is usually assumed, but where .0)) are not .
9
feasible, the quasi-experimental use of comparison groups can improve knowledge greatly if a minimum of a~~ropriately standardized and concise records is maintained.
The design examined the levels of self-esteem between two groups of inmates;
one group having been identified as varsity sports participants in one of the
Bureau's four varsity sports in the previous year, the second group being identi-
fied as non sports participants.
Each institutional activity director was requested to submit to the Planning
and Research Division a 1 ist of those inmates who participated in at least one
of the identified varsity sports programs the prior year. This I ist subsequently
identified 482 such inmate athletes who were still committed to the Bureau. The
482 inmates were numerically I isted, and from this I ist 100 inmates were randomly
selected (from a standard table of random numbers). A second group of 100 inmates
were selected to represent the non-participants. Each non-participant was
![Page 18: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/18.jpg)
individually selected so as to match with a participant, based on the variables
race, age, offense, committed county, time served and institutional location
(1 isted with respect to priority), These two groups comprised the first phase
of the project.
The second research design that the study util ized as the Randomized
Solomon Four-Group Design. Regarding the val idity of the design, this four-
group design:
... permits the control and measurement of both (1) the main effects of pretesting and (2) the ihteraction effects of pretestina and!. Furthermore, the combined effects of maturation and history can be measured if the T2 mean for group 4 is compared with the Tl means. This d~ign actually amounts to doing the experiment twice (once with pretests and once without). Consequently, if the results of the Iitwo exper iments l' are cons i stent, greater conf i dence cal1 be placed in the findings than would otherwise be possible, 13
The second phase of the study ut iIi zed se\l~ect inmate recept ions who were
initially received and processed through one of Bureau of Correction's three
diagnostic and classification centers (DCC). This phase used those inmates
who were processed in March and April of 1978.
In February 1978, the directors of the DCC's were informed of the project
and were requested to include the SEI along with the normal battery of psycho-
logical tests administered to the inmate before his subsequent prison placement
and program development. The SEI was administered to only the April commit
ments, as this would control for pre-test sensitivity. (See Figure 1.).
After ten months had elapsed (allowing sufficient time for the April 1978
commitments to become involved in anyone of the four varsity sports), the'
March and April 1978 commitments were placed into Qne of four groups (this
again corresponds to the Randomized Solomon's Four Group Design) according
10
![Page 19: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/19.jpg)
to their involvement in sports. The four groups are:
Group - Apri 1 comm i tmen t s who became involved in sports;
Group 2 - Apri 1 commitments with no involvement in sports;
Group 3 - March commitments who secome involved in sports, and
Group 4 - March commitments wi th no involvement in sports.
Chosen Level of Significance
This study selected the 0.05 level of significance at which it would test •
the hypothesis.
Traditionally, evaluation research uses the 0.05 level significance. This means that the outcome found will be attributed to change if it can be demonstrated by a test of significance that the outcome could have occured by chance 5 percent (or more) of the time. More specifically, a test of significance provides a statement of· the probabil ity that a difference as large or larger than the difference observed could have occured as the result of the operation of random phenomena - chance -when there really should be no difference - or when there should be a particular difference. These random phenomena that may produce differences include random sampl ing, random allocation of cases to treatment and control grou Pr4 and random errors of response or measurement.
Instrument ReI labil ity
No reI iabil Ity determinations have been conducted on the SEI 25 item short
11
form; however, two reI iability checks for the long form were reported in Robinson
and Shaver's Measures of SociaZ PsychologicaZ Attitudes (1973). Taylor and Reitz
(1968), authors of numerous reI iabil ity studies, have reported a .90 spl it-half
reI iabil ity for the long form, while Coopersmith displayed a test-retest
reI iabil ity of .88 over five weeks and .70 over three years.
![Page 20: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/20.jpg)
In that no accuracy determinations of the SEI short form have been
performed, the reader must conclude that the reI iabil ity is 1I ••• probably
somewhat less stable due to the shorter length. "
Instrument Validity
Coopersmith's Self-Esteem Inventory has been correlated with a number
of related self-esteem measuring scales. Comparison of Rosenberg's Self
Esteem Scale (1965) indicates that a correlation of .59 and .60 exists between
the SEI short form. A correlation of .68 was obtained from comparing the SEI
long form and Soares I Self Perception Inventories (1965). Numerous other
validity correlations may be obtained by referring to Robinson and Shaver's
Measures of Social Psychological Attitudes (1973).
Analysis of Data
This study util ized the computer center of the Pennsylvania State
University's Capital Campus in Middletown, Pennsylvania for data analysis.
Data analysis was achieved by use of the IIStatistical Package for the
Social Sciences ll (SPSS).
Study Sites
The inmate subjects used in this study were from the Pennsylvania
Bureau of Correction's ctate correctional institutions at Camp Hill, Dallas,
Graterford, Huntingdon, Pittsburgh and Rockview. The State Correctional
Institution at Muncy (Pennsylvania's female prison) was excluded from this
study since the recreational sports offered there are not of a competitive
varsity level. No subjects came from the State Regional Correctional Facil ity
at Greensburg for the reason that no inmates from that institution were
identified as a result of the random sample.
12
![Page 21: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/21.jpg)
CHAPTER IV
ANALYSIS OF DATA
Since this study is composed of two separate research designs
(a quasi-experimental design for Phase I and the Solomon1s Randomized
Four-Group Design for Phase I I), each phase will be discussed separately.
Phase I Findings
During February, 1978 the Activities Section of the Pennsylvania
Bureau of Correction identified 482 prison inmates who were labeled vnrsity
sports participants during 1977. From this 1 ist,a random sample identified
100 inmates via a table of random numbers to comprise the experimental
group. Additionally, 100 inmate non-participants were selected to comprise
the control group, based on the variables race, age, offense committed,
committing county, time served and institutional location. Of these 200
inmates identified to participate in the quasi-experimental design, 154
inmates volunteered to take the pretest. The frequency distribution for the
154 inmate1s SEI score can be found on Table 1. Viewed separately in groups,
the frequency distributions for the two groups (experimental group, i.e.
sports participants, N=84; and control group, i.e., non-participants, N=70)
can be found on Tables 2 and 3, respectively.
Initially, the difference-of-means test would probably be preferred
by most researchers, but since the study did not util ize two independent
Ii.: .. i
![Page 22: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/22.jpg)
SEI Score
06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean Median Mode
19.117 20.038 22.00
TABLE I
Self-Esteem Inventory Responses
for Phase I
Frequency
1 2 1 1 2 4 3 4 5 7 4 7
10 19 13 17 25 13 10 6
1511
Std. Err 0.337 Std. Dev. 4.182 Range 19.00
Relative Frequency (percent)
0.6 1.3 0.6 0.6 1.3 2.6 1.9 2.6 3.2 4.5 2.6 4.5 6.5
12.3 8.4
11.0 16.2 8.4 6.5 3.9
100.0
Cumulative Frequency (percent)
0.6 1.9 2.6 3.2 4.5 7. 1 9.1
11.7 14.9 19.5 22. 1 26.6 33.1 45.5 53.9 64.9 81.2 89.6 96. 1
100.0
100.0
t~ in i mum 06.00 Maximum 25.00 Variance 17.489
14
::' ; , ,
![Page 23: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/23.jpg)
SEI Score
TABLE 2
Self-Esteem Inventory Responses
for Sports'Partf6ipants
Frequency
Re lat ive Frequency (percent)
Cumulative Frequency (percent)
~ 15 n
![Page 24: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/24.jpg)
SEI Score
07 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean Median Mode
18.5]11 19.500 22.00
TABLE 3
Self~Esteem Inve~tory Responses
for Non-Participants
Frequency
2 1 1 3 3 2 2 4 1 4 5 7 6 7
11 3 5 3
70
Minimum Max [rHum Rang-e
7.00 25.00 18.00
Relative Frequency (percent)
2.9 1.4 1.4 4.3 4.3 2.9 2.9 5.7 1.4 5.7 7.1
10.0 8.6
10.0 15.7 4.3 7.1 4.3
100.0
Variance Std. Err. Std. Dev.
Cumulative Frequency (percent)
2.9 4.3 5.7
10.0 14.3 17. 1 20.0 25.7 27.1 32.9 40.0 50.0 58.6 68.6 84.3 88.6 95.7
100.0
100.0
20.601 0.542 4.539
16
![Page 25: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/25.jpg)
sampling techniques, the t test is technically inappropriate. However,
since the two groups did not util ize a direct pair-by-pair comparison, a
t test was util ized.
A significance level of O.OS and a one-tailed test were selected to
define the significance levels in the first phase.
The data analysis (which was done manually) in Phase I demonstrated
that there was a statistically significant difference found between the
sample and cohort means (t=I.88, df=lS2, p=.OS). This statistical signi-
ficance supported the hypothesis that those inmates who participate in an
organized varsity sports program will tend to have a higher level of self-
esteem than those who do not participate in sports programs.
"Significant" here does not mean "important" or "consequence"; it is used here to mean "indicative ofll or '.'signi§yingll a true difference between the two populations.
However, since this researcher had access to computer facil ities at
Pennsylvania State University's Capital Campus in Middletown, Pennsylvania,
the data was re-examined util izing the Statistical Package for the Social
Sciences (SPSS), which ",., is an integrated system of computer programs
designed for the analysis of social science data. lll6 This data analysis
produced findings (t=l.6448, df=lS2, p=.OS) that alter the findings as
reported in "Sports Program Evaluation: Phase I,ll
This difference of opinion can be attributed to the more sophisticated
level of information which was obtained from the computer than what was avail-
able manually.
Specifically, the area of disagreement focuses on the difference found
17
between the calculated t values of the two methods of data analyses. In Phase I,
![Page 26: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/26.jpg)
18
in order to support the research hypothesis, the calculated t value had to
exceed the critical value of t~ i.e. 1.6450. When the data was analyzed
manually, the caiculated t'value was 1.88, which exceeded the already defined
critical value of t and hence supported the hypothesis. However, the
calculated t value was defined at 1.6448 when relying upon the data analysis
from the SPSS program. Sinc~ this value does not exceed the critical t value
of 1.645, the research hypothesis cannot be supported! A manual check'of Phase
I's data was done using the analysis as found in the SPSS program, and the
manual check supported the computer findings.
Therefore, Phase I data cannot support the research hypothesis that
there is a statistically significant different found between sports
participants and non-participants in regard to self-esteem.
The relatively small difference between the calculated t values (0.003)
does allow the researcher to hypothesize that he has made an incorrect decision
based on the measurement of the samples and not on the populations.
This incorrect decision is known as a Type II Error.
Since no significant differences among the sample and cohort means were found the null hypothesis was accepted. Conssquently, there is a possibility of a Type I I Error .•.. Furthermore, since inferential statistics is concerned with inferences from samples and since population characteristics are rarely known'l,he researcher seldom knows if a Type I I Error has occurred.
Phase I I Findin~s
Phase II requires a complex research design which can accommodate
four groups (April sports participants and non-participants as well as
Marchhs sports participants and non~particlpants): two groups receiving
treatment i.e., partictpatl~n in the organized sports program; two groups
![Page 27: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/27.jpg)
~--,-~,-,,---''''''. -.,.-~.,..,-------------..,....------------------~-~
19
receiving a pretest; and all four groups receiving a posttest. The research
design selected for use in Phase I I was consequently identified as the
Randomized Solomon Four-Group Design (see Figure 1).
The Randomized Solomon Four-Group Design lends itself to a multitude
While this design addresses the methodological perimeters of this study, I of statistical analyses, however these analyses require that specific conditions
be met prior to the acceptance of any research conclusions.
the imposed requirements makes it difficult, if not impossible, for a controlled
experiment such as this to take place within a prison. While the study did not
alter the construction of the design, it did have to modify the requirements.
The primary condition which this design had to ignore (which will subse-
quently inval idate the "findings" throughout the entire study) was the require-
ment that each of the four groups be randomly assigned. This condition was
unfortunately the central controlling factor for the design, as II ••• the random
assignment of subjects makes it possible to assume that the pretest scores for
groups 3 and 4 would have been simil iar to the pretest scores attained by
groups and 2." 18
The random assignment of inmates was not possible since the study could
not require that only select inmates participate in organized varsity sports
programming. Hence, April receptions who ten months later were identified as
varsity sports participants were placed in group 1, while the remaining April,
1978 receptions who were not identified as sports participants were placed in
group 2. likewise, March, 1978 receptions who were identified as sports
participants were placed in group 3, with the remaining March, 1978 receptions
placed in group 4. This placement by sports participation negates the possibility
.. ,,':,-------------------_._---------------------
![Page 28: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/28.jpg)
~I " 20
FIGURE 1.
Randomized Solomon Four-Group Design
*1 Tl = pretest with the Self-Esteem Inventory
*2 X = independent variable, i,e., sports participation
*3 T2 = posttested with the Self-Esteem inventory
*4 A'l ' prl commitments
*5 March commitments
I j
![Page 29: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/29.jpg)
. '
FIGURE 2.
Randomized Solomon ' s Four-Group Design
for Phase I I Receptions
Treatment Group Pretested (sports involvement) Posttested
T1 X 12
2 T1 89
3 X 11
4 97
Total 218 209
218 - April receptions who were pretested (Groups 1 and 2) 193 - March receptions who were identified (Groups 3 and 4) 4TT - Total number of inmates e1 igib1e for posttesting (Groups 1-4)
21
'1 __ I~ --.-,---------~---------------------------. '
![Page 30: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/30.jpg)
of assuming that both sports participants and non-participants eminate from
the identical population.
The issue of random assignment to one of the four groups can be
appreciated by the reader when the number of cases per group is examined.
Under the random assignment of subjects to the four groups, an equal number
of cases per group would be the obvious advantage, which would subsequently
allow the researcher to util ize the Z test, which is particularly appropriate
to determine if there exists a significant difference between two sample means,
especially those of large samples, i.e., a sample size greater than 30. The t
test is satisfactory for large samples, though with the t test if the groups
in question have an equal number of results, " ... a violation of the assumption
of equal v'ariances does not affect the validity of the t test. Furthermore
the t " ... assumes the two samples come from two populations with equal means
and equal variances.,d9
For example, only 18 of the 218 inmates who were received by the Bureau
in April, 1978 were later identified as sports participants, and only 12 out
22
of the 18 volunteered to take the posttest. Likewise, of the 200 non-participants,
only 89 elected to participate in the posttest. This vast difference in group
size (N of group = 12; and N of group 2=89) enables the reader to appreciate
why the findings of this study are technically inval id. For example, on Table 5,
the frequency distributions for those inmates completing the SEI pretest can be
found. Since there were only 12 subjects completing the pretest, the effect
that a low SEI score, say, 04, could have a dramatic impact simply because of
the low number of subjects in the group. The addition of a SEi score of 04
in Table 5 would cause the mean score to decrease 1.2 points; while in Tabl.e 7
the addition of an SEI score of 04 would only decrease their group's mean SEl
score by . 158.
I;.'·· ~i
i
~
t
![Page 31: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/31.jpg)
SEI Score
Mean Median Mode
18.500 19.00 17.00
TABLE 4 Self~Esteem Inventory Responses of
Phase I I IS Group # 1
(April sports participants)
Frequency
Minimum Maximum Range
Pre-test
13.00 22.00
9.00
Relative Frequency (percent)
Cumulative Frequency (percent)
Variance 8.273 Std!. Err. 0.830 Std. Dev. 2.876
23
~. I ,
![Page 32: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/32.jpg)
SE1 Score
11 13 14 18 21 22 23 24
Total
Mean Median Mode
19.500 21. 167 21.00
TABLE 5
Self-Esteem Inventory Responses of
Phase II I S Group #1
(April Sports Participants)
Frequency
1 1 1 1 3 1 3 1
12
Posttest
Min imum 11 .00 Maximum 24.00 Range 13.00
Relative Frequency (percent)
8.3 8.3 8.3 8.3
25.0 8.3
25.0 8.3
100.0
Var i ance Std. Err. Std. Dev.
Cumulative Frequency (percent)
8.3 16.7 25.0 33.3 58.3 66.7 91.7
100.0 100.0
19.727 1.282 4.442
24
![Page 33: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/33.jpg)
------------ - - --~---- - ---- -- -------
Pretest SEI Score
05 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean Median Mode
16.876 17.429 20.00
TABLE 6
Se1f~Esteem tnventory Responses of
Phase tt~s Group # 2
CApri 1 non ..... parHcipants)
Frequency
1 5 1 4 6 2 3 6 7 3 7 . 7 5 8 6 6 6 4 2
89
Minimum Maximum Range
5.00 25.00 20.00
Relative Frequency (percent)
1.. 1 5.6 1.1 4.5 6.7 2.2 3.4 6.7 7.9 3.4 7·9 7.9 5.6 9.0 6.7 6.7 6.7 4.5 2.2
100.0
Variance Std. Err. Std. Dev.
Cumulative Frequency (percent)
1.1 6.7 7.9
12.4 19. 1 21.3 24.7 31.5 39.3 42.7 50.6 58.4 64.0 73.0 79.8 86.5 93.3 97.8
100.0
100.0
23.769 0.517 4.875
25
![Page 34: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/34.jpg)
TABLE 7 Self-Esteem Inventory Responses of
Phase I I ts Group # 2
(April non-participants)
26
![Page 35: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/35.jpg)
Posttest SEI Score
11 14 15 18 19 20 23 24
Total
Mean 19.091 Median 19.250 Mode 19.00
TABLE 8 Self-Esteem Inventory Responses of
Phase I I IS Group # 3
(March sports participants)
Frequency
1 1 1 1 2 1 2 2
11
Minimum Maximum Range
11.00 24.00 13.00
Relative Frequency (percent)
9. 1 9. 1 9. 1 9.1
18.2 9. 1
18.2 18.2
100.0
Variance Std. Err. Std. Dev.
Cumulative Frequency (percent)
9. 1 18.2 27.3 36.4 54.5 63.6 81.8
100.0
100.0
18.891 1.310 4.346
27
'.
![Page 36: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/36.jpg)
;'
Posttest SEI Score
09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean Median Mode
18.031 18.050 18.00
TABLE 9 Self-Esteem Inventory Responses of
Phase I I IS Group # 4 (March non-participants)
Frequency
1 3 1 4 5 5 7 8 9
10 7 8 5 9 7 7 1
97
Minimum Maximum Range
9.00 25.00 16.00
Relative Frequency (percent)
1.0 3. 1 1.0 4.1 5.2 5.2 7.2 8.2 9.3
10.3 7.2 8.2 5.2 9.3 7.2 7.2 1.0
100.0
Variance Std. Err. Std. Dev.
Cumulative Frequency (percent)
1.0 4.1 5.2 9.3
14.4 19.6 26.8 35. 1 44.3 54.6 61.9 70. 1 75.3 84.5 91.8 99.0
100.0
100.0
15.405 0.399 3.925
28
------------------------------------------------------------------------
".\
"
,"
i,',,:, .:
t ;,
![Page 37: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/37.jpg)
I
29
Furthermore, analysis of variance cannot be comput'ed because the
assignment of the number of cases to the groups was not independent of each
other, as those inmates who were not identified as varsity sports partici-
pants were automatically dumped in group 2. However, to control for con-
sistency (actually to use unrel iable statistical techniques consistently
throughout the study), a t test will be employed.
Group Analysis
When the pretest SEI Scores for group 1 (April receptions/sports partici-
pants) were compared against group 2 (April receptions/non-participants) there
did not exist a significant difference between the two sample means (18.S00 and
16.876 for groups l' and 2, respecthtely; t~L 1'2, df=99, p=.OS). The analysis
found on Tables 4 and 6 represent the frequency distributions for group 1 and
2 1 s pretest, respectively; 1 ikewise, Tables Sand 7 contain the frequency
distributions for the posttests of groups 1 and 2, respectively. Thus, the
pretest supports the argument that there is no statistical significance between
the self-esteem of inmates prior to their involvement or non-involvement in
organized varsity sports.
Upon examining Table la, the reader can see that both groups 1 and 2 did
improve their SEI scores over the course of the 10 months following their recep-
tion. Group 1 increased their SEI pretest mean score of 18.soo to 19.500, how-
ever this 1.00 increase did not display a statistical significance (t=.6S, df=22 ,
p=.OS). However, group 2 did increase their mean SEI score a total of 1.27
(from 16.876 to 18.146), which did create a statistically difference between the
two means (t=1.8080, df=176, p=.OS).
Finally, when the posttests for groups 1 and 2 were compa:ed, a mean
difference of 1 .3S4 was displayed. However, this mean difference between the
two samples did not display a statistically significant difference (t=.97, df=99 ,
p=.OS).
"jl. " ,
![Page 38: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/38.jpg)
I-·-··~
Group
A~ril partici~ants
White ...... " ........... , Black ••••••••••••••• eo •••
For entire group .........
A~ri i non-~artic~ants
White .................... Bl ack ...........•........ For entire gro~p ........•
March Participants
Wh i te ................... ~ Black .................... For entire group .•.•.....
March non-~articipants
Wh i tel .. , .............. <II •
Black ....... , ........ ,q ••
For entire group .........
TABLE 10
Breakdown of the Self-Esteem Inventory Scores by
Race and Group
SEI SEI SEI SEI pretest pretest pos ttes 1:" posttest x score variance x score variance
17.250 13.583 15.550 33.667 19. 125 5.839 21.500 2.857 18.500 8.273 19.500 19.727
15.955 29.672 18.273 19.970 17.778 16.859 18.022 21.022 16.876 23.769 18. ]l~6 20.285
23.5 0.500 18. 11 17.611 19.091 18.891
18.218 16.285 17.786 14.514 18.031 15.405
Mean Age
27.7 23.2 24.7
26.2 27.0 26.6
27.0 26.1 26.2
26.9 25.9 26.5
N
4 '8
12
44 45 89
2 9
11
55 42 97
w o
![Page 39: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/39.jpg)
In conclusion, with regards to groups 1 and 2, the only statistically
significant increase in mean sample SEI scores occurred in group 2 when their
pr"etest SEL score of 16.876 increased to 18.146, as was displayed on the post
test. Therefore, while the two groups did not differ significantly when
comparing pre and posttests separately, group 1 did maintain a mathmatically
higher SEI score during the two testing periods.
Since groups 3 and 4 were not pretested, it is impossible to determine
what, if any, effects sports participation had upon group 3. The comparison of
posttests between groups 3 and 4 displayed a 1.06 difference between their
sample mean SEI scores, however this difference was not sufficient enough to
produce a statistically significant difference (t=1.26, df=106, p=.05).
31
The design also permits the researcher to determine what effects (if any)
sports participation, pretesting and outside fnfluences had upon the group
means. Overall, the participation in the Bureau IS organized varsity sports
program can account for a .564 mean SEI Score difference between groups; the
pretest alone accounted for only a 0.11550 mean SEI score difference between
groups; and the combined effects of unnamed influences or other variables
accounted for a 1.155 mean SEI score difference. Clearly, sports participation
can not lay claim to the improvem6nt of an inmate's self-esteem. Most probably,
the mathmatical increase in the SEI scores would have to be attributed to other
outside variables, such as acceptance of self, internal ization of self-concept,
etc.
When the study was broken down to review the frequency distributions of
sports participants (N=23) versus non-participants (N=186) (see Tables 11 and 12,
respectively), analysis produced no stati~tical significance between the post
test SEI mean scores for groups 1 and 2 (19.304 and 18.086, respectively), with
t=1.3l, df=207, p=.05.
![Page 40: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/40.jpg)
SEI Score
11 13 14 15 18 19 20 21 22 23 24
Total
Mean 19.304 Median 20.667 Mode 23.000
TABLE 11
Self-Esteem Inventory Responses of
Posttests in Phase I I for
Varsity Sports Participants
Posttest
Relative Frequency
Frequency (p~rcent)
2 8.7 1 4.3 2 8.7 1 4.3 2 8.7 2 8.7 1 4.3 3 13.0 1 4.3 5 21.7 3 13,0
23 100.0
Minimum 11 .000 Maximum 24.000 Range 13.000
32
Cumulative Frequency (percent) ~
8.7 13.0 21.7 26. 1 34.8 43.5 47.8 60.9 66.2 87.0
100.0
100.0
Variance 18.494 Std. Err. 0.89' Std. Dev. 4.3")0
![Page 41: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/41.jpg)
. "
SEI Score
4 8 9
10 1 1 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean Median Mode
18.086 18.500 20.000
TABLE -'12
Self-Esteem Inventory Responses of
Posttests in Phase I I for
Non-Participants
Pasttest
Frequency
1 1 4 4 4 5 9
10 12 14 12 17 15 20 11 17 12 15 3
186
Minimum 4.000 Maximum 25.000 Range 21.000
Relative Frequency (percent)
0.5 0.5 2.2 2.2 2.2 2.7 4.8 5.4 6.5 7.5 6.5 9.1 8.1
10.8 5.9 9.1 6.5 8.1 1.6
100.0
Cumulative Frequency (percent)
0.5 1.1 3.2 5.4 7.5
10.2 15. 1 20.4 26.9 34.4 40.9 50.0 58.1 68.8 74.7 83.9 90.3 98.4
100.0
100.0
Variance 17.647 Std. Err.' 0:308 Std. Dev. 4.201
33
, -
1-
'
"
" t,
![Page 42: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/42.jpg)
· " When the posttest SEI scores were compared with regard to race (see
Tables 13 and 14), virtually no difference at all was discovered when
examining the mean score, median score, the standard error and standard
deviation.
In examining the SEI scores by race and group, white sports partici
pants (N=6) were compared to black participants (N=17), though there resulted
no statistica·1 significance between the mean sample scores. With regard to
the non-participants, when the while (N=99) and black (N=97) part'icipants
were compared to each other, again no statistical significance resulted.
34
---- ~------ --------------------~~~~~~~--'!
![Page 43: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/43.jpg)
· .
SEI Score (Posttest)
4 8 9
11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean 18.238 Median 18.818 Mode 20.000
TABLE 13
Self-Esteem Inventory Responses of
Posttests in Phase I! for Whites
Whites
Relative Frequency
Frequency (percent)
1 1.0 1 1.0 2 1.9 3 2.9 4 3.B 5 4.8· 5 4.8 6 5.7 8 7.6 5 4.8 9 8.6
11 10.5 12 11.4 4 3.8 9 8.6 9 8.6 9 8.6 2 1.9
105 100.0
Minimum 4.000 Maximum 25.000 Range 21.000
35
Cumulative Frequency (percent)
1.0 1.9 3.8 6.7
10.5 15.2 20.0 25.7 33.3 38.1 46.7 57.1 68.6 72.4 81.0 89.5 98. 1
100.0
100.0
Variance 18.510 Std. Err. 0.420 Std. Dev. 4.302
![Page 44: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/44.jpg)
· .
SEI Score (Posttest)
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Total
Mean 18.202 Median 18.500 Mode 18.000
TABLE ·14
Self-Esteem Inventory Responses of
Posttests in Phase I I for Blacks
Non-Whites
Relative Frequency
Frequency (percent)
2 1.9 4 3.8 3 2.9 1 1.0 5 4.8 7 6.7 7 6.7 6 5.8 7 6.7
10 9.6 6 5.8 9 8.7
10 9.6 9 8.7 8 7.7 9 8.7 1 1.0
104 100.0
Minimum 9.000 Maximum 25.000 Range 16.000
36
Cumulative Frequency (percent)
1.9 5.8 8.7 9.6
14.4 21.2 27.9 33.7 40.4 50.0 55.8 64.4 74.0 82.7 90.4 99.0
100.0
100.0
Variance 17.250 Std. Err. 0.407 Std. Dev. 4.153
![Page 45: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/45.jpg)
·"
•
CHAPTER 5
CONCLUSIONS
This study was obviously hampered by a multitude of problems, several
of them being the following: (1) the inabil ity to randomly assign inmates
to a specific group in the Randomized Solomon Four-Group Design, (2) the
inabil ity to control for the vast difference in group sizes when comparing
sports participants versus non-participants, (3) the inabil ity to use a set
statistical devise to determine the difference-of-the means between small
and large groups, (4) the relatively short time span that this study had to
be conducted over, and (5) the extreme loss of cases over ~ short period of
time.
Other researchers, noteably Martinson et al , have found that correctional
research is hampered by a host of problems. In addressing the evaluation studies
of correctional treatment and the relatively few num~er of evaluative studies
that meet acceptable guideline criteria, Martinson 'concluded that ..•
In part, this is a function of the difficulties involved in measuring attitude change, including (1) the loss of cases over time, (2) the 1 ikel ihood that as time passes, test scores within a group will become more alike, (3) the general instability of the measuring devises utilized, which makes it impossible to determine whether a respondent's change in attitude or personal ity is due to the unrel iabil tty of the test or to actual change, and (4) the inabil ity or the failure to take into account the fact that computing group average change scores overlooks the "room-for-change" individuals may have. (For example, an individual who scores well at time one has less IJroomlJ to improve than an individual who scores poorly at time one.
The analysis of data has indicated that there was no significant difference
between sports participant and non-participant. While this may be true for the
![Page 46: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/46.jpg)
• • •
inmate population, it is equally worth noting that the inmate athlete still
$ possessed a higher mean SEI score than did the non-athlete. What the state
prison system did, then, was to make available to newly received inmates who
already possessed an more than adequate self-esteem a means of not only
preserving that self-esteem but improving on it. Clearly it is ~vident that
the Bureau's organized s~orts program is more attractable to the inmate who
requires athletic participation so that he may identify himself with either
a particular sport or a recognizable group of inmates .
..
38
![Page 47: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/47.jpg)
..
FOOTNOTES
1. Lewis E. Lawes, Twenty Thousand Years at Sing Sing (Ray Long and Richard R. Smith, Inc., 1932), p. 218.
2. Elmer H. Johnson, Crime, Correction, and Society (The Dorsey Press, 1968), p. 589.
3. Paul WI Tappan, Crime Justice and Correction (McGraw-Hill Book Company, Inc., 1960), p. 696.
4. Douglas Lipton, Robert Martinson, and Judith Wilks, The Effectiveness of Correctional Treatment: A Survey of Treatment Evaluation Studies (P raeger Pub 1 i shers, 1975), p. 332.
5. ~,p. 295.
6. Ibid, p. 3.
7. Walter E. Schaefer, IIParticipation in Interscholastic Athletics and Del inquency: A Prel iminary Study, "Social Problem~, XVII (1969), 41.
8. I bid, p. 42
9. James B. Nolan, "Athletics and Juveni Ie Del inquency," Sociolo51Y of Education, 28 (1954/?5) , p. 265.
10. John P. Robinson and Phill ip R. Shaver, Measures of Social Psychological Attitudes (The Institute for Social Research, The University of Michigan), p. 84.
11. Daniel Glaser, "Five Practical Research Suggestions for Correctional Administrators, "Crime and Del inquency, XVII (January, 1971), p. 39.
12. I bid.
13. Solomon Kobrum, Inferential Statistics (McGraw-Hill Book Company, Inc., 1974), p. 4.
14. Martinson, Effectiveness of Correctional Treatment, p. 19.
15. Norman H. Nie, et al. Statistical Packa51e For the Social Sciences (New York: McGraw-Hill Book Company, 1975), p. 267.
16. ~., p. XXI.
17. Schuyler W. Huck, William H. Cormier, and Will iam G. Bounds, Jr., Readin51 Statistics and Research (New York: Harper & Row, 1974), p.45.
I
![Page 48: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/48.jpg)
, 40
•
18. Kobrum, Inferential Statistics, p. 41.
19. ~li., p. 134.
20. Martinson, Correctional Treatment, p. 621.
![Page 49: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/49.jpg)
..
..
" , . BIBLIOGRAPHY
Glaser, Daniel. "Five Practical Research Suggestions for Correctional Administrators," Crime and Del inquency, XVII (January, 1971), 32-40.
Johnson, Elmer Hubert. Crime, Correction, and Society revised ed. Homewood, 111.: The Dorsey Press, 1968.
Lawes, Lewis E. Twenty Thousand Years in Sin~ Sing. New York: Long & Smith, Inc., 1932.
Lipton, Douglas; Martinson, Robert; and Wilks, Judith The Effectivesness of Correctional Treatment, New York: Praeger ~ub1 ishers, 1975.
Nolan, James B. "Ath1etics and Juvenile Delinquency," Sociology of Education, 28 (1954-1955), 263-65.
Robinson, John P. and Shaver, Phil ip R. Measures of Social Psychological Attitudes, Ann Arbor, Michigan: The University of Michigan, 1973.
Schafer, Walter E. "Participation in Interscholastic Athletics and Del inquency: A Prel iminary Study," Social Problems, 17 (1969), 40-47.
Solomon, Kobrum. Inferential Statistics. New York: McGraw-Hill Book Company, Inc. 1974.
Tappan, Paul W. Crime,Justice and Correction. New York: McGraw-Hill Book Company, Inc.) 1960.
liThe Fighting Con," Newsweek, May 1, 1972, p. 60.
Wicker, Tom. A Time To Die. New York: Quadrangle/The New York Times Book Co., 1975 .
·
'1.:.;1
r " It , i, f
![Page 50: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/50.jpg)
-" .
• ..
..
APPENDICES
"'
![Page 51: DocumentV>](https://reader031.vdocuments.us/reader031/viewer/2022022121/6212c10cfd1d1f41ad5f90a9/html5/thumbnails/51.jpg)
, I
•
..
I
I • . r . I .. [
APPEND!X A
P NP Age Name
Sb Fb Bb Bx Race BC # SCI@
March Apri'l --SELF-ESTEEM INVENTORY
Please read the follo\·ling statements and check (V) whether the statement is I ike you or unl!ke you. There are no right or wrong answers.
1. often wish I were someone else. I ike me
2. find it very hard to talk in front of a group. I ike me
3. There are lots of things about myself I'd change if I could. I ike me
4. I can make up my mind without too much trouble. I ike me
5. I'm a lot of fun to be wi tho I ike me
6. I get upset eas i Iy at home. I ike me
7. It takes me a long time to get 'lsed to anything new. I ike me
8. I'm popular with people my own age. I ike me
9. My fami Iy expects too much of me. I ike me
10. My fami Iy usually considers my feelings. I ike me
11. give in very eas i I y. I ike me
12. It's pretty tough to be me. I ike me
13. Things are all mixed up in my life. I ike me
14. Other people usually follow my ideas. I ike me
IS. I have a low opinion of myself. I ike me
16. There are many times when I felt like leaving home. I ike me
17. I often feel upset about the work that I do. I ike me
18. I'm not as nice looking as most people. I ike me
19. If I have something to say. I usually say it. I ike me
20. My family understands me. I ike me
21. Most people are better liked than I am. I ike me
22. usually feel as if my family is pushing me. I ike me
23. often get discouraged at what am d~ing. I ike me
24. Things usually don't bother me. I ike me
25. I can't be depended on. I ike me
43 i
--
unlike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unlike me
unl ike me'
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
unl ike me
un like me
unl ike me
unl ike me
un like me
unl ike me